We previously identified decreased bone density as a complication of anorexia nervosa. There are serious implications to this finding. Since the majority of patients are at an age when bone mass is normally increasing, peak bone mass may be compromised. The many anorectics who remain ill for years or who do not recover could be at jeopardy for de- veloping debilitating bone disease relatively early in life. Even recovered anorectics could be at increased risk of eventually developing bone disease later in life from reduced bone mineral stores. Understanding of bone density problems in anorexia is hampered by the absence of long term follow-up data and the limited knowledge of normal bone development in the adolescent/young adult.age group. Much current information on normal development derives fran cross sectional rather than longitudinal study. In the proposed longitudinal investigation bone density changes in anorexia nervosa will be documented over a five year period across a range of treatment outcomes. A group of matched normal controls also will be followed. Seventy-five subjects with anorexia nervosa drawn from the departmental inpatient service and 80 carefully screened normal controls recruited through public notice will be entered. Bone density will be evaluated by means of dual photonabsorpticmetry of the lumbar spine and single photon absorptiometry of the non-dominant mid-radius. Baseline values will be obtained including serum pregnancy test, estradiol, 25-OH-Vit D, Ca, inorganic PO4, alk phosphatase, T4, T3Ru, TSH, PTH, BUN, creatinine, TIBC, total protein and 24 hour urine cortisol and creatinine. Activity and food logs together with the Eating Attitudes Test will be completed. Anorectic subjects will repeat the evaluation every six months and controls will repeat evaluations with serum pregnancy tests but without other blood or urine studies annually. Length of follow-up will vary depending on time of subject entry. Follow-up period will range from 24 to 54 months with two-thirds of both groups having at least three years follow-up. Hypotheses will be tested concerning between group differences related to duration of illness, menstrual status, and weight. Hypotheses will be tested related to within group differences regarding weight, menstrual status, exercise, vitamin D, Ca, cortisol, and estradiol. Results will permit better assessment of the risk of osteoporosis in anorexia nervosa and the extent to which recovery of bone density is possible. Findings may have value in the more general understanding of bone disease in the adolescent/young adult age group. Control data will provide much needed information on the normal course of bone density development.